A Study on the Implementation of PSI Protocol Using Perfect Hashing Based on OKVS as an Advanced Pseudonymization Technique
Year: 2024
Type: article
Abstract: OKVS (Oblivious Key-Value Store) is a data structure that enhances the security of traditional Key-Value Stores by incorporating the Oblivious Key property.This property ensures that the data can only be accessed when a search query is performed, which significantly enhances security and privacy.This feature is particularly useful for enhancing security and privacy, which is crucial in strengthening data protection in environments such as My Data and cloud computing.Techniques such as Bloom Filter and Cuckoo Hashing are utilized to propose efficient search and access methods.However, Bloom Filter, known for its fast lookup speed and efficient memory usage, is a probabilistic data structure with the disadvantage of potential false positives, where non-existent elements might be mistakenly identified as present.Cuckoo Hashing, on the other hand, employs two hash functions and two hash tables to avoid hash collisions by moving the new key to the second hash table to avoid conflicts.Despite its advantages, Cuckoo Hashing can lead to infinite loops under certain conditions.To address these drawbacks, the proposed PSI (Private Set Intersection) protocol leverages Perfect Hashing within the OKVS framework, ensuring collision-free operations.By using OKVS, which enhances security over traditional KVS data structures, and proposing Perfect Hashing to avoid collisions, we enable the safe implementation of PSI, particularly in data environments that demand accurate search and retrieval, such as those containing sensitive medical and financial data.
Author: Ji Yeon Lee
Institution: Halla University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4402931397